The genome.ucsc.edu web site provides the primary point of access to the reference human genome sequence for many tens of thousands of scientists and medical researchers worldwide. It is used over the World Wide Web by more than 5,000 scientists each day, servicing more than 150,000 requests for information. The site includes human gene sequences, their location in the genome, expression levels in different tissues, alternatively spliced transcripts, and protein products. Users get quick links to other databases. There is an integrated view that provides information on each gene's function in normal and diseased states. The site supports 31 genomes in addition to human, and produces multiple genome alignments and other comparative genomics analysis. This analysis, along with other large scale data such as ChlP/CHIP, sheds light on regulatory as well as coding regions. [unreadable] [unreadable] Expansions to the project are needed to handle the flood of additional genomes and other large data sets, to further develop comparative genomics, to improve the gene sets, to increase interactions with other databases, and to add additional support for transgenic studies in model organisms. We also propose three new projects to help link together genomics and more medically oriented research: (1) a suite of tools for working with SNP association studies and other genomic scans, (2) tools to support medical sequencing and identify causative mutations, and (3) support for new representations of human polymorphism both in the germ line and in cancerous and aging cells. By helping scientists tie changes in the genome to medically relevant phenotypes, these new projects will greatly facilitate the next generation of disease studies. [unreadable] [unreadable] Relevance: At least half of all diseases have a substantial genomic component. This work will help scientists better understand these diseases, and develop new treatments. [unreadable] [unreadable] [unreadable] [unreadable]